Evaluation of Pedestrian-Perceived Comfort on Urban Streets Using Multi-Source Data: A Case Study in Nanjing, China

Urban street comfort is a crucial measure of street environmental quality. However, traditional evaluations primarily focus on physical elements, often neglecting pedestrian perceptions. In this study, considering five core evaluation dimensions—safety, mobility, aesthetics, perceptibility, and convenience—an innovative quantitative evaluation model is proposed to assess pedestrian-perceived comfort on urban streets by integrating physical environmental factors and subjective experiences. This analysis comprises two steps: evaluation indicator extraction and weight application. Indicators are extracted from multi-source data (street-view images, real-time traffic data, points of interest, and pedestrian surveys) using a deep learning method. A comprehensive weighting method combining entropy weight and the analytic hierarchy process is used to determine the relative importance of each factor. This study focuses on Nanjing as a case study, and the results reveal significant variations across the five dimensions and their 11 secondary indicators. Street environment safety (0.143) is critical for street safety, while the degree of street traffic congestion (0.121) dominates street mobility. Street aesthetics is primarily influenced by building enclosure (0.105), and street convenience is strongly affected by the number of surrounding bus stops (0.260). Spatial analysis indicates higher comfort levels in urban centers due to well-developed infrastructure, whereas peripheral areas face challenges from inadequate facilities. Notably, areas around parks demonstrate elevated pedestrian-perceived comfort levels, highlighting the importance of green spaces. Overall, the proposed evaluation system provides new insights from the perspective of pedestrian experience and offers valuable guidance for urban planning and policy.